Understanding the loss surface of a neural network is fundamentally important to the understanding of deep learning. This paper presents how piecewise linear activation functions substantially shape the loss surfaces of neural networks. We first prove that {\it the loss surfaces of many neural networks have infinite spurious local minima} which are defined as the local minima with higher empirical risks than the global minima. Our result demonstrates that the networks with piecewise linear activations possess substantial differences to the well-studied linear neural networks. This result holds for any neural network with arbitrary depth and arbitrary piecewise linear activation functions (excluding linear functions) under most loss functions in practice. Essentially, the underlying assumptions are consistent with most practical circumstances where the output layer is narrower than any hidden layer. In addition, the loss surface of a neural network with piecewise linear activations is partitioned into multiple smooth and multilinear cells by nondifferentiable boundaries. The constructed spurious local minima are concentrated in one cell as a valley: they are connected with each other by a continuous path, on which empirical risk is invariant. Further for one-hidden-layer networks, we prove that all local minima in a cell constitute an equivalence class; they are concentrated in a valley; and they are all global minima in the cell.

Dineo Lioma's energy and enthusiasm are infectious, and the way she speaks about things such as artificial intelligence (AI), DNA, enzymes and gene sequencing to aid those in the field of healthcare makes one believe there is hope for our virus-stricken world. Lioma has co-founded two medical technology companies, founded a third and currently runs two of them – and she is not yet 30 years old. She has a master's degree in micro and nanotechnology enterprise with distinction from Cambridge University in the UK. Cambridge offered Lioma the chance to do her PhD there to "work out how to harness solar and mechanical energy", but the engineer, who grew up fiddling with electrical plugs in her family's home in Bloemfontein, Free State, and obtained her BSc in metallurgical and materials engineering with 24 distinctions from Wits University, declined the offer. "I knew I wanted to work in the field of health and to help South Africa progress. There was not a lot going on in micro and nanotechnology here, so I came home. I wanted to give back," she says.

Neanderthals fed regularly on mussels, fish and other omega-3-rich marine life including seals, which likely impacted their cognitive abilities, a new study claims. Archaeological digs along the Portuguese coast reveal the evidence that our cavemen ancestors had as much fondness for seafood as modern humans today. Both Neanderthals and early Homo sapiens tucked into'surf and turf', from molluscs, crabs, fish, waterfowl and dolphins to horse, goat and red deer, as well as pine nuts. The findings are based on ancient remains in the cave of Figueira Brava, Portugal, dating to roughly 106,000-86,000 years ago – when Neanderthals settled in Europe. Figueira Brava is 18.6 miles (30km) south of Lisbon on the slopes of the Serra da Arrábida, a natural park facing south, about a 45-minute drive from Lisbon'Pretty much every potential source of food that existed in the environment they [Neanderthals] exploited and used it,' said Professor João Zilhão, an expert in palaeolithic archaeology at the University of Barcelona.

Chest X-ray is the first imaging technique that plays an important role in the diagnosis of COVID-19 disease. Due to the high availability of large-scale annotated image datasets, great success has been achieved using convolutional neural networks (CNN s) for image recognition and classification. However, due to the limited availability of annotated medical images, the classification of medical images remains the biggest challenge in medical diagnosis. Thanks to transfer learning, an effective mechanism that can provide a promising solution by transferring knowledge from generic object recognition tasks to domain-specific tasks. In this paper, we validate and adopt our previously developed CNN, called Decompose, Transfer, and Compose (DeTraC), for the classification of COVID-19 chest X-ray images. DeTraC can deal with any irregularities in the image dataset by investigating its class boundaries using a class decomposition mechanism. The experimental results showed the capability of DeTraC in the detection of COVID-19 cases from a comprehensive image dataset collected from several hospitals around the world. High accuracy of 95.12% (with a sensitivity of 97.91%, a specificity of 91.87%, and a precision of 93.36%) was achieved by DeTraC in the detection of COVID-19 X-ray images from normal, and severe acute respiratory syndrome cases.

This paper addresses the difficulty of forecasting multiple financial time series (TS) conjointly using deep neural networks (DNN). We investigate whether DNN-based models could forecast these TS more efficiently by learning their representation directly. To this end, we make use of the dynamic factor graph (DFG) from that we enhance by proposing a novel variable-length attention-based mechanism to render it memory-augmented. Using this mechanism, we propose an unsupervised DNN architecture for multivariate TS forecasting that allows to learn and take advantage of the relationships between these TS. We test our model on two datasets covering 19 years of investment funds activities. Our experimental results show that our proposed approach outperforms significantly typical DNN-based and statistical models at forecasting their 21-day price trajectory.

Deep Convolutional Neural Networks (DCNNs) is currently the method of choice both for generative, as well as for discriminative learning in computer vision and machine learning. The success of DCNNs can be attributed to the careful selection of their building blocks (e.g., residual blocks, rectifiers, sophisticated normalization schemes, to mention but a few). In this paper, we propose $\Pi$-Nets, a new class of DCNNs. $\Pi$-Nets are polynomial neural networks, i.e., the output is a high-order polynomial of the input. $\Pi$-Nets can be implemented using special kind of skip connections and their parameters can be represented via high-order tensors. We empirically demonstrate that $\Pi$-Nets have better representation power than standard DCNNs and they even produce good results without the use of non-linear activation functions in a large battery of tasks and signals, i.e., images, graphs, and audio. When used in conjunction with activation functions, $\Pi$-Nets produce state-of-the-art results in challenging tasks, such as image generation. Lastly, our framework elucidates why recent generative models, such as StyleGAN, improve upon their predecessors, e.g., ProGAN.

Edge intelligence refers to a set of connected systems and devices for data collection, caching, processing, and analysis in locations close to where data is captured based on artificial intelligence. The aim of edge intelligence is to enhance the quality and speed of data processing and protect the privacy and security of the data. Although recently emerged, spanning the period from 2011 to now, this field of research has shown explosive growth over the past five years. In this paper, we present a thorough and comprehensive survey on the literature surrounding edge intelligence. We first identify four fundamental components of edge intelligence, namely edge caching, edge training, edge inference, and edge offloading, based on theoretical and practical results pertaining to proposed and deployed systems. We then aim for a systematic classification of the state of the solutions by examining research results and observations for each of the four components and present a taxonomy that includes practical problems, adopted techniques, and application goals. For each category, we elaborate, compare and analyse the literature from the perspectives of adopted techniques, objectives, performance, advantages and drawbacks, etc. This survey article provides a comprehensive introduction to edge intelligence and its application areas. In addition, we summarise the development of the emerging research field and the current state-of-the-art and discuss the important open issues and possible theoretical and technical solutions.

Hisham Elamir is a data scientist with expertise in machine learning, deep learning, and statistics. He currently lives and works in Cairo, Egypt. In his work projects, he faces challenges ranging from natural language processing (NLP), behavioral analysis, and machine learning to distributed processing. He is very passionate about his job and always tries to stay updated about the latest developments in data science technologies, attending meet-ups, conferences, and other events.

As we have announced recently, Appsilon Data Science's AI for Good initiative is working together with biodiversity conservationists at the National Parks Agency in Gabon and in collaboration with experts from the University of Stirling. Part of our role in the project is to develop an image classification algorithm capable of classifying wildlife seen in images taken by camera traps located in the forests of Gabon. The project has received support from the Google for Education fund which allowed us to embark on this journey with the immense power of the latest computational resources at hand. Below are some interesting findings we made so far. Stay tuned for more news on the progress! We have recently participated (and taken the 5th place out of 811 participants) in the Hakuna Ma-data competition, in which we classified images of wildlife from the savannahs of Serengeti.

The companies already partner on charging system operations managed services. The new deal, which is already being implemented, includes network operations center services and field services in radio, core and transmission technology. AI-driven intelligent and data driven operations are part of the deal. AI-related efficiency benefits, automation and data analytics will enable MTN Benin and Ericsson to run predictive operations to enable a shift from reactive to proactive IT and network management – boosting customers' experiences, as well as network quality and performance. Stephen Blewett, Chief Executive Officer, MTN Benin, says: "Network managed operations play a significant role in improving MTN customers' satisfaction and enhancing customer experience as well as enabling revenue growth and cost efficiency. We expect advanced technologies like AI, automation and analytics to further accelerate operational transformation through this new managed services partnership."